⚛️ quantum physics

On the emergence of quantum mechanics from stochastic processes

This paper generalizes the stochastic-quantum correspondence by establishing a dictionary between stochastic kernels and CPTP maps, demonstrating that the emergence of quantum dynamics and phase information from a phase-blind stochastic description arises as a memory effect when the lifted family satisfies Chapman-Kolmogorov divisibility, thereby admitting a Lindblad master equation form.

Original authors: Jason Doukas

Published 2026-03-27

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Jason Doukas

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are watching a complex movie. You see actors moving, lights changing, and scenes unfolding. Now, imagine someone tells you: "Actually, this movie isn't real. It's just a series of random coin flips happening on a giant, invisible board."

This is the bold idea Jason Doukas explores in his paper: Can the weird, magical world of Quantum Mechanics be explained as just a fancy kind of random gambling?

Usually, we think of Quantum Mechanics (QM) and Classical Probability (like rolling dice) as completely different languages.

Classical Probability: You have a bag of red and blue marbles. You pick one. It's either red or blue. The math is straightforward.
Quantum Mechanics: The marble is in a "superposition" (both red and blue at once) until you look. It can interfere with itself like a wave. It has "phases" (a hidden internal clock) that classical marbles don't have.

Doukas argues that QM isn't a totally different universe. Instead, it's a compressed, hidden version of a very complex random process.

Here is the breakdown using everyday analogies:

1. The "Shadow" vs. The "Object"

Imagine you are in a dark room with a 3D sculpture. You shine a flashlight on it, and you see a 2D shadow on the wall.

The Shadow: This is the Stochastic Process (the random probabilities). It only shows you "what happened" (e.g., the marble landed on Red). It has no depth; it's "phase-blind."
The Object: This is Quantum Mechanics. It has depth, hidden angles, and internal structure (the "phase").

Doukas asks: Can we reconstruct the 3D object just by looking at the 2D shadow?
His answer is yes, but with a catch. You can't just look at the shadow of one step. You have to look at the history of the shadow.

2. The "Memory" in the Machine

In a normal game of dice, if you roll a 6, the next roll doesn't care that you rolled a 6 before. That's a "Markovian" process (no memory).
In Quantum Mechanics, things do remember. If you do a trick, then another trick, the result depends on the order and the history, even if the individual steps look random.

The Analogy:
Imagine a magician (the Quantum System) and a scribe (the Stochastic Process).

The scribe only writes down the final trick: "The rabbit appeared."
The magician, however, did a complex sequence of sleight-of-hand moves involving hidden cards and mirrors (the Phases).
If the scribe only writes down "Rabbit," they lose the magic.
Doukas's Insight: The "magic" (the phase) is actually just compressed memory. The scribe's notes are incomplete because they only look at the immediate past. But if you look at the entire story of how the rabbit got there, the "magic" is just the hidden path the magician took.

3. The "Lift" (Going from Flat to 3D)

The paper introduces a mathematical tool called a "Lift."

Think of the stochastic process as a flat map of a city. It tells you the distance between two points.
The "Lift" is like taking that flat map and folding it into a 3D origami sculpture.
Suddenly, the "flat" distances on the map make sense because they are actually the result of the sculpture's 3D shape.

Doukas shows that for any random process, you can build this 3D "Quantum" version.

The Catch: If the random process is "indivisible" (meaning you can't break it down into simple, independent steps without losing information), the 3D sculpture needs extra room to store that missing information.
In Quantum Mechanics, that "extra room" is the Off-Diagonal Phase. It's the hidden storage space where the history of the process is kept, invisible to the simple probability map but crucial for the next step.

4. Why Do We See "Weird" Quantum Stuff?

You might ask: "If it's just random, why do we see interference and entanglement?"

The Analogy:
Imagine two runners, Alice and Bob.

Alice runs a path where she takes a left turn.
Bob runs a path where he takes a right turn.
If you only look at where they start and where they finish, they look identical.
But if you ask them to run a second lap immediately after, Alice's left turn puts her in a position to take a shortcut, while Bob's right turn blocks him.
The difference wasn't in the first lap; it was in the hidden memory of the first lap.

In Quantum Mechanics, the "Phase" is that hidden memory. When we measure a quantum particle, we are essentially asking, "What path did you take?" The "weirdness" (interference) happens because the particle is remembering its entire history, not just its current location.

5. The "Division" Event (Measurement)

The paper also explains what happens when we "measure" something.

In the quantum world, measuring is like resetting the game.
Before you measure, the system is carrying a heavy backpack of "memory" (phases, history).
When you measure (look at the result), you force the system to drop the backpack. You force it to become a simple, flat probability again.
This is called a "Division Event." It's the moment the complex, memory-filled quantum story collapses into a simple, random fact.

The Big Takeaway

Jason Doukas is saying: Quantum Mechanics isn't magic; it's just a very efficient way of storing history.

Classical Probability is like a diary that only records the daily weather.
Quantum Mechanics is like a diary that records the weather plus the hidden reasons why the weather changed (wind patterns, pressure systems).
The "Phases" in quantum mechanics are just the hidden notes in the diary that explain why the next day's weather is different from what a simple guess would predict.

By understanding this, we see that the "weird" quantum world is actually just a compressed version of a complex, memory-filled random process. The "spooky" action at a distance and the "magic" of superposition are just the result of a system remembering its past in a way that a simple probability map can't show.

1. Problem Statement

The paper addresses the foundational tension between classical stochastic processes and quantum mechanics. While quantum mechanics is traditionally viewed as distinct due to its reliance on complex amplitudes, phase interference, and unitary evolution, the stochastic–quantum correspondence (proposed by Barandes et al.) suggests that quantum dynamics can be derived from underlying stochastic processes on a configuration space.

However, the original formulation of this correspondence faces several unresolved issues:

Restrictive Structure: The original "θ-process" formulation excludes standard continuous-time Markov chains (CTMCs) because it requires transition kernels to be $O(dt^2)$ near the identity, effectively forbidding first-order leakage ( $R \neq 0$ ).
Indivisibility: It is unclear how non-Markovian, indivisible stochastic processes (which violate the Chapman-Kolmogorov composition law) can consistently map to quantum dynamics without ad-hoc assumptions.
The Phase Problem: How can a "phase-blind" stochastic description (containing only probabilities) generate the rich, phase-sensitive interference effects characteristic of quantum mechanics?
Observer Dependence: The role of measurement, state collapse, and the distinction between trajectory realization and physical intervention needs clarification within this framework.

2. Methodology

The author generalizes the stochastic–quantum correspondence by moving beyond the specific $\theta$ -representation to a broader framework involving Completely Positive Trace-Preserving (CPTP) maps (Kraus maps).

Lifting Framework: The paper defines a "lift" of an arbitrary finite-dimensional stochastic kernel $\Gamma(t \leftarrow s)$ to a linear map $\phi_{t,s}$ on the space of bounded operators $\mathcal{B}(\mathcal{H})$ .
Compatibility Condition: A lift is valid if it satisfies a commutative diagram condition: projecting the lifted map's output onto the diagonal (configuration basis) must reproduce the original stochastic evolution.
$\Pi(\phi_{t,s}(J(p(s)))) = J(\Gamma(t \leftarrow s)p(s))$
where $J$ embeds probability vectors into diagonal density matrices, and $\Pi$ projects operators onto their diagonal.
Explicit Construction: The author constructs explicit CPTP lifts for any stochastic kernel using Kraus operators, showing that the transition kernel elements are sums of squared moduli of Kraus matrix elements:
$\Gamma_{ji} = \sum_\beta |K_{\beta, ji}|^2$
Divisibility Analysis: The paper investigates the Chapman-Kolmogorov (CK) divisibility of the lifted family. It posits that while the stochastic kernel $\Gamma$ may be indivisible (non-Markovian), the lifted quantum map $\phi$ may admit a CK-composable structure.
Memory Compression: The analysis explores how higher-order stochastic memory (multi-time conditional probabilities) is compressed into the off-diagonal (phase) degrees of freedom of the lifted density operator.

3. Key Contributions

A. Generalization to Arbitrary Stochastic Kernels

The paper proves that any stochastic transition kernel $\Gamma$ (not just $\theta$ -processes) can be lifted to a CPTP map. This resolves the limitation of the original formulation, allowing for the inclusion of classical CTMCs and processes with genuine first-order leakage ( $R \neq 0$ ) within the quantum correspondence framework.

B. The Role of CK Divisibility

The author identifies CK-divisibility of the lifted family as the decisive constraint.

If a stochastic process admits a CPTP lift that satisfies the CK composition law ( $\phi_{t,s} = \phi_{t,u} \circ \phi_{u,s}$ ), the lift admits a Lindblad master equation form.
This establishes a direct link: Indivisible stochastic processes (non-Markovian) correspond to open quantum systems governed by Lindblad dynamics, where the "memory" of the stochastic process is encoded in the non-commutative phase structure of the quantum state.

C. Resolution of the Phase-Blind Paradox

The paper explains how phase information emerges from phase-blind data:

Path-Space Memory: A single-step stochastic kernel $\Gamma$ does not uniquely determine the process; it under-determines the multi-time joint probabilities.
Compression: The "missing" information required to distinguish between stochastically indistinguishable unitary evolutions (e.g., $U_X$ vs. $U_Y$ where $|U_X|^2 = |U_Y|^2$ ) resides in the higher-order conditional structure (history dependence) of the stochastic process.
Operationalization: When the process is lifted to $\mathcal{B}(\mathcal{H})$ , this history dependence is compressed into the off-diagonal elements (phases) of the density matrix. These phases are invisible in single-step statistics but become operationally relevant through composition and system-apparatus coupling, reproducing basis-dependent quantum statistics.

D. Divisibility Criterion (Theorem 1)

The paper provides a rigorous mathematical criterion for when a stochastic process admits a consistent quantum lift that respects divisibility:

Theorem: A stochastic process is "C-divisible" (admits a stochastic composition law) at time $t_1$ if it is "Q-divisible" (admits a CK-composable CPTP lift) at $t_1$ and the lifted state $\rho(t_1)$ remains diagonal for all initially diagonal states.
This clarifies the relationship between measurement-induced "division events" (screening off history) and the existence of a valid quantum description.

4. Key Results

Existence of CPTP Lifts: Every finite-dimensional stochastic kernel $\Gamma$ admits at least one CPTP lift. The original $\theta$ -process is a special case (rank-1 Kraus representation) that restricts the dynamics to be effectively unitary or trivial in the short-time limit.
Lindblad Emergence: The requirement that the lifted family be CK-divisible forces the dynamics to take the Gorini-Kossakowski-Sudarshan-Lindblad (GKSL) form. Thus, open quantum system dynamics is the natural "lifted" expression of general stochastic dynamics.
Recovery of CTMCs: Classical CTMCs are recovered as the restriction of Lindblad dynamics to the diagonal subalgebra, provided the initial state and measurements are diagonal. The paper shows that standard CTMCs can be embedded into this framework, contrary to the original $\theta$ -process restriction.
Memory vs. Phase: The paper demonstrates that the space of path-space memories (stochastic multi-time laws) is exponentially larger than the space of unitary lifts. The quantum lift acts as a compression mechanism, storing the complex history dependence of the stochastic process in the compact phase degrees of freedom of the density matrix.
Measurement and Division: The paper distinguishes between "trajectory realization" (conditioning on a fact) and "intervention" (changing the effective law). It proves that a division event (where history is screened off) is the necessary condition for defining a history-independent linear propagator and, consequently, a valid quantum lift.

5. Significance

Unification of Frameworks: The paper successfully unifies classical stochastic processes and quantum mechanics under a single mathematical umbrella. It suggests that quantum mechanics is not a fundamentally different theory but a specific, highly structured sector of stochastic dynamics where history dependence is compressed into phase.
Clarification of Quantum Features: It demystifies quantum phenomena like interference and entanglement, reinterpreting them as manifestations of memory effects in an underlying stochastic process that are inaccessible to single-step probability observations but revealed through composition and measurement.
Foundational Insight: By removing the "observer" from the center of the dynamical update (replacing it with physical interventions and division events), the paper offers a more objective, realist interpretation of quantum state updates.
Technical Advancement: The derivation of the divisibility criterion and the explicit dictionary between stochastic kernels and Kraus maps provide new tools for analyzing open quantum systems and non-Markovian stochastic processes.

In conclusion, Doukas argues that the "quantumness" of the world arises not from a rejection of stochasticity, but from the specific way in which stochastic memory is structured and compressed into the phase degrees of freedom of a lifted operator algebra. The correspondence reframes the foundational question: rather than asking how quantum mechanics emerges from randomness, it asks why our fundamental theories appear to lie in the specific "unistochastic" regime where first-order leakage is suppressed or structured in a way that admits a CK-divisible quantum lift.